Grok is one of the most mind-bending technologies we've come across. But it's only powerful if you use it for the right things. 11 ways to use Grok and save you hours of time: (🔖Bookmark for later)

Full open source reproduction of DeepSeek R1 in progress ⏳

The trend in robot manipulation is to collect data and learn end-to-end policies. Do we still need perception or state estimation? My analogy is how Tesla and Waymo are building self-driving cars. The car can understand the 3D world around it. Robots need to have this capability

open source at work

Built Secure Conversational AI Agent based on Deep-Seek-R1 🚀 which trying to focus on Data Privacy Driven Shared Over chat experience with end-to-end encryption & secure conversation history. github.com/Agentic-OS/Sec… Agent

Andrej Karpathy Bytedance open sourced their RL framework for LLMs recently, plan to dive into it: github.com/volcengine/verl

I read up on DeepSeek’s learning algo, GRPO. GRPO: group relative policy optimization How GRPO works: 1 • model generates a group of answers 2 • compute score for each answer 3 • compute avg score for entire group 4 • compare each answer score to avg score 5 • reinforce

Another big week of AI and Robotics news. So, I summarized everything announced by Booster Robotics, Adobe, OpenAI, Figure, ByteDance, Google, Perplexity, Apptronik, Humanoid, Mentee Robotics, and more Here's everything you need to know and how to make sense out of it:

The TypeScript team just announced that they are porting to Go. It's already 10x faster 🤯

Awesome LLM Apps is trending at #1 on GitHub with 25k stars. It features 50+ step-by-step tutorials on: - AI Agents and Multi-agents - RAG (Agentic and Local) - MCP AI Agents - LLM Apps with memory - & so much more 100% free and Opensource.

new blog post "There Are No New Ideas In AI.... Only New Datasets" in which i summarize LLMs in exactly four breakthroughs and explain why it was really *data* all along that mattered... not algorithms

Holy sh*t… Anthropic just released a prompt engineering bible. It’s free. It’s insanely well written. And it covers real Claude use cases. Every AI builder should bookmark this. Here’s what’s inside + link:

FUTURE WORK – direct extraction we're working on directly extracting training data from models using RL and other methods. we'll be presenting our first work on this in COLM, and expect more in this space we may be able to directly extract data from the 120b model.. one day 😎

Your brain doesn’t erase memories — it just loses the keys that unlock them. This paper introduces a key-value memory system that splits how the brain stores and retrieves. Keys go to the hippocampus for quick access. Values go to the neocortex for high-fidelity storage.

One of the most beautiful blogs I have come across in a while

A bit late, but finally got around to posting the recorded and edited lecture videos for the **How to AI (Almost) Anything** course I taught at MIT in spring 2025. Youtube playlist: youtube.com/watch?v=0MYt0u… Course website and materials: mit-mi.github.io/how2ai-course/… Today's AI can be

🏓🤖 Our humanoid robot can now rally over 100 consecutive shots against a human in real table tennis — fully autonomous, sub-second reaction, human-like strikes.

Cancel your wedding if you have to but watch this video on how to use LLMs. It covers a wide array of usecase from - basics of prompting - using deep research - NotebookLM youtu.be/EWvNQjAaOHw?si…

probably the cleanest backend playlist to exist

One of the best math books I've ever read: MIT's "Mathematics for Computer Science" Its writing style is brilliant, and it covers everything: - Linear algebra - Series - Logic - Probability - Number theory - Graphs You can find the PDF here: